Mage Theory Crafting: Combat Ratings - 4.01
jak3676 - Tue, 10/19/2010 - 10:01
in
Level 60 70 80 81 82 83 84 85
Spell Hit 8 12.615 26.23 34.445 45.232 59.42 78.022 102.446
Crit 14 22.077 45.906 60.278 79.156 103.986 136.538 179.28
Haste 10 15.769 32.79 43.056 53.539 74.275 97.527 128.06
Mastery N/A N/A 45.906 60.278 79.156 103.986 136.538 179.28
At Level 80:
Mage Int: 1 Int provides 0.006 crit
Mage Int: 166.67 Int provides 1% crit
Level Hit Cap % Hit Rating Required @ Lvl 80
0 4% 105
+1 5% 131
+2 6% 158
+3 17% 446 (skull level boss)
At Level 85:
Mage Int: 1 Int provides 0.001541 crit
Mage Int: 648.91 Int provides 1% crit
Level Hit Cap % Hit Rating Required @ Lvl 85
0 4% 410
+1 5% 512
+2 6% 615
+3 17% 1742 (skull level boss)
Keep in mind that none of the Mage specs have any abilities that give us talented hit% anymore.
I'm also told that Boomkins and Shadow Priests lost their 3% hit rating raid buff (boss debuff
technically). So I imagine I'm not the only mage trying to gear from 8% hit (formerly +6% talented
and 3% raid buff) all the way up to 17%.
*edit - I originally posted this under the Frost Mage thread, but it's not spec dependant so I moved it.
| Posts: | |
User login
Navigation
Mage Theory Crafting 101: What the stats are and what they do
Mage Theory Crafting 101: What the stats are and what they do
I am writing this from a Mage perspective, but a lot of this will be pretty class independent – but there are small differences.
Health & Mana: I won’t go into what these are – if you need help with that, this thread will be way over your head. I do need to say though that these stats do not directly impact anything else in the game. (e.g. just because you have something happens that increases your mana or health directly, does not mean you will do more damage or have more mana regeneration – however things like Evocation and Mage Armor work off a % of your total mana so these abilities do benefit from a larger mana pool).
Strength: Increases our attack power, but we should never, ever physically attack anything with the intention of doing damage. So don’t go there. (Seriously – if you run out of mana then use your wand to shoot stuff – don’t ever try to melee anything.) If you see anything with strength on it, it isn’t for you.
Agility: Just as meaningless as Strength is for mages – if you see something with lots of haste and crit on it, but it has agility, then it’s meant for some other class.
Stamina: Provides 1 health per stamina for the first 20 points of stamina, and 10 health per point of stamina thereafter. Different races have minor differences in starting stamina, but even at level 1 the difference is small.
Intellect: Increases Spell power, mana points and a small % to critical strike. Each point of intellect gives player characters 1 point of spell power, 15 mana points (except from the first 20 points of Intellect that provide 1 mana for each point instead). Int's impact on crit is addressed in the Crit section. Mages do have a passive Wizardry talent that increases our Int by 5%. So something that has 100 int will give us 105 spell power, 1575 mana and a percentage more Crit depending on our level (see charts).
Spirit/Mana Regen/Combat Regen: As of 4.01 this is another meaning stat for Mages. This impacts a character’s mana regeneration when not in combat – but that’s essentially meaningless to us. If we need to regenerate mana when not in combat we can drink – we’re mages just conjure up some water or strudel. We don’t care about mana regeneration out of combat.
We no longer have any abilities that convert Spirit into Crit or anything that that will change our in combat regeneration based on our out of combat regeneration. (Mage Armor now does its combat regeneration based on our total mana pool, not a percentage of our out of combat regeneration.) So if you see a nice piece of cloth with spirit on it – leave it for the priests. There’s no arguing about it anymore, if it has spirit on it, its healy gear, not dps gear. Combat regeneration is a flat 5% of your base mana pool per five seconds on the character sheet. This is not modified by gear, spec, or buffs. Otherwise Mages only regain mana in combat by using potions, mana gems, evocation or switching to Mage Armor. The general play style is that a mage will enter combat full on mana and (depending on the length of combat) use up that mana over the course of the fight and then pause to drink before entering the next round of combat.
Spell Power: At the introductory level, lets just say that more Spell Power = more damage. The exact formula for how much it increases the base ability for each spell is different depending on the coefficient (we’ll leave that for the 201 or 301 level of theory crafting). For the most part we get SP from Int on a 1:1 ratio, but we do still find spell power as a separate stat on caster weapons, a variety of procs in the game and enchants. More precisely we have Wizardry (passive talent for 5% more Int), Arcane Brilliance (6% increase of spell power) and in raid situation we should have a varity of buffs for more Int and more SP. This essentially means we get 1.113 spell power per point of raw Int when self buffed and ~16.5 mana.
Haste and Haste Rating: The ability to increase casting speed and decreases the time for the global cooldown (gcd) after casting a spell. Haste Rating is the amount required to gain 1% increased haste - this amount increases based on level of the caster. The speed increase is applied similarly for cast time and channelled spells. Up until the point where you start hitting a "soft cap" more haste is always better if you have the mana to support it (It’s not possible to hit the “hard cap” where ALL of your spells would be cast in less time than the gcd.)
Haste Rating stacks additively with itself but Haste stacks multiplicatively. That means that if you have 328 haste rating at level 80, you will have +10% haste, no matter how many different sources and items that haste rating comes from. If you then used something like Bloodlust for +30% haste, you would have (1 + 10%) * (1 + 30%) = 1 + 43% haste. If you used both troll Beserking and Bloodlust together you would have (1 + 10%) * (1 + 20%) * (1 + 30%) = 1 + 71.2% haste. The formula to determine your final casting speed is: Resulting Cast Time = Base Cast Time / [(1 + Item Haste Rating) * (1 + Haste source 1) * (1 + Haste source 2) * ... * (1 + Haste source n)]. Many people incorrectly assume that haste reduces your cast time, it doesn't. Haste increases cast speed. The difference may be subtle, but it's important for proper theory crafting. If you have 50% haste and a spell that takes 10 seconds to cast - that would not be: 10 seconds * 50% haste = 5 second cast. Correctly calculated it would be: 10 seconds / (1 + 50%) = 6.667 seconds.
Haste results in an increase to damage per second (DPS) – i.e. with 50% haste you will do 50% more damage within period of time. But haste does not increase your damage per mana (DPM) - so within that period of time you will do 50% more damage, but you will also use 50% more mana. The general global cooldown on all spells is 1.5 secs. Spell haste can reduce the global cooldown for spells, to a minimum of 1 second. To reduce the global cooldown to 1 second, it would take 50% haste or a haste rating of 1639.5 (at level 80). This is often referred to as the "soft cap" for haste, but for most practical purposes it is valuable to go far beyond this until the point where your most commonly cast spell has a cast time of 1.0 seconds under the most optimal conditions (i.e. simultaneously using/proc'ing all sources of haste and haste rating). Keep in mind that there are many different sources of Haste Rating and Haste in the game (e.g. gear, trinket proc's, bloodlust, enchantments, potions, etc.) It is possible to reduce the cast time of a spell to less than 1.0 seconds, but then you will be limited by the gcd anyway so any haste/haste rating that would take you below 1.0 seconds is essentially wasted.
Hit Chance/Hit Rating: The amount of +hit needed to guarantee that you'll hit a target 100% of the time. If someone has enough Hit Chance or Hit Rating to guarantee that will hit 100% of the time, they are said to be “Hit Capped”.
The amount of Hit Chance needed is determined based on level difference between you and your target. Depending on your level it takes an ever increasing amount of Hit Rating to equal a given Hit Chance %. There is a different % for PvE vs PvP targets. For all classes and specs, Hit Rating is the most important stat to max first, dps wise. You can essentially think of a 1% chance to miss as 1% less damage AND 1% less chance to crit AND 1% more mana per point of damage AND a 1% less chance of proc’ing your various abilities that come based on hitting a target. Get as close to the cap as possible, but keep in mind that any points over the cap are completely wasted. We no longer have any talents or raid buffs that decrease the amount of hit needed – those were pre 4.0 abilities that have since been removed from the game.
level difference PVE hit cap PVP hit cap -4 0% 0% -3 1% 1% -2 2% 2% -1 3% 3% 0 4% 4%(PvP Hit Cap) +1 5% 5% +2 6% 6%(Dungeon Hit Cap) +3 17% 13%(Raid Boss Hit Cap) +4 28% 20% +5 39% 27%
Crititcal Chance (Crit) and Crit Rating: The ability to increase the chance of scoring a critical hit. Crit Rating is the amount required to gain 1% increased crit - this amount is based on the level of the caster.
For a level 80 character: Crit Chance % = ((Intellect / 166.6667) + Class Constant) + (Crit Rating / 45.91). Class constants are 1.85 for Druids, 3.6 for Hunters, 0.91 for Mages, 3.336 for Paladins, 1.24 for Priests, 2.2 for Shaman, and 1.701 for Warlocks. To see the resulting Crit Chance %, you can just look at your character sheet in-game. Talents that affect the crit chance of all spells will be included in this number, Talents that only increase the crit chance of a specific spell will not be shown here - you have to add that % manually.
Spell critical strikes deal 150% normal damage without talents. Some spellcasting classes have talents to increase this Crit Multiplier; however, most of the crit multiplier talents were removed from the game in 4.0, but the format is still the same. (Spell critical heals will only heal for 150% of their normal amount.)
Unlike Haste and Haste Rating, Crit and Crit Rating are only additive, not multiplicative when determining the Crit Multiplier. There is one important exception - the "diamond" meta gems (Chaotic Skyflare Diamond, Chaotic Skyfire Diamond, Relentless Earthstorm Diamond, and Relentless Earthsiege Diamond). For Mages we should always be using the Chaotic Skyflare Diamond (CSD). The tooltip says that this increases your crit damage by 3%, but this is before taking other talented crit damage increases. By itself the CSD increases your Crit multiplier from 1.5 to 1.545 – this can be further multiplied by Ignite for Mages.
Crit Muliplier = (1 + (150% * CSD) * (1 + [sum of talented crit damage increases]))) * [damage proc'd from critical strikes]
Example: Fireball with ignite (40% extra damage after crit proc), CSD 3% multiplicative crit damage, and 4pc T7 gear (5% crit damage)
Crit Multiplier = 1 + ( (150% * CSD) - 1) * (1 + [4pc T7]) ) * [ignite]
Crit Multiplier = 1 + ( (150% * 1.03) - 1) * (1 + 5% ) ) * 1.4 = 227.115%
Note that the crit multipliers for 4.0 are much lower than in all previous expansions. Previously there were a variety talents to increase critical strike damage – taken together Some mages approached 350% more damage on critical strikes.
Mastery and Mastery Rating: This is the new stat introduced with 4.0. Its exact impact depends on your chosen specialization, but Mastery Rating is similar to the other rating stats in that it takes an ever increases amount to increase by a give % based on the level of the caster. The value of Mastery and Mastery Rating will vary by class and spec, but at preliminary amounts available by reforging at level 80, it looks to be slightly less valuable than Hit Rating, but more valuable than Haste or Crit Rating.
Spell Penetration and Resistance: Spell Pen is a stat only currently useful to PvP to counteract the amount of resistance another player may have (typically 65 points from the Druid/Paladin buff). Non-Player Characters (PvE mobs) do not generally have resistance that can be mitigated with Spell Penetration so raiding mages will avoid this stat. At level 80 it takes about 50 points of resistance to mitigate 10% spell damage – this will increase as we level from 80-85.
Resilience: Resilience is used in PvP to mitigate damage done by other players. At level 80 it takes about 35 points of resilience to mitigate 1% of the damage done by other players and their pets – it is not applied to damage done by Non-Player Characters (PvE mobs) so raiding mages will avoid this stat.
Armor, Dodge, Parry, Block: We do gain a small % of dodge as we level and with the 4.0 revisions our armor has a decent amount of ability to reduce physical damage, but essentially we don’t have any way to improve our ability to mitigate or avoid incoming damage that hits us – it’s our job to make sure we’re standing in a safe place to avoid incoming damage. Mages should never intentionally try to increase these stats at the expense of damage dealing stats.
Mage Theory Crafting 201: Coefficients and Spell Power
Spell
*post 4.01 nerfs
Mage Theory Crafting 201: Coefficients and Spell Power
As much as I hate to go back in time to revisit "the way things used to be"; I think it's necessary to understand what's going on now on why the theory crafting community is having such a problem putting together good 4.01 calculators.
In the ol' days we used this:
Time cast divided by 3.5 will yield how much +heal or + spell damage will happen.
For example, if u have +150 to spell damage and cast a 2 second spell, to calculate + spell damage you would take 2/3.5 and then that would be the proportion of the 150+ damage that would be added.
2/3.5*150=86 damage added.
Blizzard put this system in place so that if you got an item with spell power on it (a rare stat back in the ol days) the +spell damage wouldn't impact shorter duration spells more than it impacted longer cast spell. Under this formula something that increased your spell power by 100 would proportionally affect something with a 2.0 second cast time the same as a 3.0 second cast time.
The first issues with this formula were that not every spell followed this formula. There were a few talents that decreased cast time - so did the spell take the new reduced cast time or th orignal cast time? It was kinda hit or miss, but we figured it out. Then there were a few odd balls like Frostbolt which had a base cast time of 2.5 seconds used a 3.0/3.5 coefficient. Eventually we came to understand that Blizz assigned a .5/3.5 coeffecient tax "because the spell had a build in snare". (Keep in mind that Blizz never actually publishes any of this information - its just lots of fans spending hours (OK, days) hitting target dummies and anything else gave us consistant results and then spending hours pouring over logs and spreadsheets to figure out what it all means.) But at the end of day we "knew" that it was just a matter of figureing out what the x/3.5 nubmer was.
So now we get to the table above (pulled from EJ). First you'll notice that these aren't exactly whole numbers anymore. You can look at a few like Arcane Missles, Fire Blast and Frostbolt and assume that the poster pretty much got them right. The numbers are pretty close to rounded to the nearest 1/2 and they have some resemblance to cast time too. But then we get things like FrostFire Bolt, Ice Lance and Fireball that just don't fit the X/3.5 format. Did Blizz drop some of code and now some spells use X/2.5 or just some other X/Y format? Is there a completely new mechanic here we haven't figured out yet? Who knows?
For now lets just assume that the spell power coefficents posted above are correct as stated (even if they don't exactly fit the cast time/3.5 model). Using this formula, the spell power coefficient table and the spells base damage you should now be able to figure out the correct damage you can expect on the target dummy.
Spell damage = spell base damage * spell power coefficient * available spell power.
Historcially "spell base damage" has been reported accurately on the spells tooltip in your spell book. We are still seeing several numbers change from week to week since the 4.01 patch. This could be impacting our ability to determine accurate coefficients.
This is generally the point where someone pipes up and something like "OK, OK - I'm not really paying attention to you anymore. I just want to know how much spell power I need to do 1% more damage so I can start comparing its value to Crit, Haste and Mastery." Well basic algebra says that if you want 1% more spell damage on the left side of the equation (assuming spell base damage and the spell power coefficient stay the same), you need to have 1% more spell power on the right hand side. For those who already aren't following - you need 1% more spell power than you have now to gain 1% more spell damage - it's not a set number.
You can look at a T10 geared mage and assume they'll have about 3000 spell power when raid buffed so 1% more damage will be about 30 spell power. A T10.5 geared mage has about 3300 spell power when raid buffed so 1% more damage is up 33 spell power at that point. For those looking ahead to cata, best estimate is that we'll have about 6k sp when we get our dungeon blues at level 85, and we'll get over 10k by the end of T11 content.
For our purposes, something like this is probably close enough
Gear Level Gear Score Spell Power
T10.5 6k 3.3k
T10 5.5k 3k
T9-9.5 5k 2.5k
T8-8.5 4k 2.k
T7-7.5 3k 1.5k
I spec AGI for ranged crit.
I spec AGI for ranged crit.
P.S.
Great post, Mage. :) I'm always looking for new information to give me that extra edge in DPS. Thanks for the info.
Mage Theory Crafting 301: Calculating a spells total DPS
I typed 301 out somewhere (probably at work seeing as I can't find it at home). So this is just a place holder for now
Mage Theory Crafting 401: Putting it all together
I need to dig this one up too. I could've sworn I posted already, but I guess not.
Mage Theory Crafting for dummies
Download RAWR, follow it's advice like you life depends on it (except when it F's up and tells you to do something stupid - what can I say its a beta).
http://rawr.codeplex.com/
Seriously - I'll give kudos to anyone who actually reads 1/2 of what I type. I do think its critical for good dps to actually know what their talents and stats do. But when it comes right down to it, you really just want to know if item X is better than item Y or if you should really reforge/regem/re-enchant to trade some stat for some other stat. As much as I'm all about spreadsheets, recount and analyzing world of logs, I always have RAWR open on another monitor any time I'm rolling on loot. They don't always get everything perfect, but when it comes to mage dps, RAWR gets more scrutiny by the worlds theorycrafting leaders than any other tool.
One other tool I do want to mention is Magegraf. It's not so much of a true similator (like RAWR), but more of a set of well put together formulae. It's goal is to achieve the same result, but while RAWR attempts to re-create what your dps should look like given any given set of inputs, Magegraf is much simplier. Here's what Vontre came up with for total stat weighting, assuming Best-In-Slot T11 gear.
Magegraf RESULTS, T11
RAWR stat weighting (iLevel 343)
I just ran my char through the RAWR optimiser for a frost mage (average item level = 343 currently) and it gave me the following stat weights (normalized for SP). Of note is that it had me reforge everything possible into Hit and Crit. It actually got me to the Frost mage Crit cap (33.33% with all the raid buffs), but just short of the Hit cap (17%).
SP: 1.00
Int: 1.28
Crit: 0.25
Hit: 0.82
Haste: 0.47
Mastery: 0.30
For the fun of it, I turned off one of the raid buffs for crit (leaving me 5% below the crit cap). That boosted the stat weighting for crit quite a bit, as expected, but of course it also bosted the weight for Int a bit as it gives a small amount of crit as well. It also lowered the value of mastery as we'll have fewing things shattering because we're not crit caped. This is pretty similar to the Magegraf numbers above, with the differences generally due to different gearing.
SP: 1.00
Int: 1.33
Crit: 0.54
Hit: 0.82
Haste: 0.47
Mastery: 0.29
This really goes to prove the general stat weighting for Frost mages is: Int >> SP >> Hit below 17% >> Crit below 33% > Haste > Mastery > Crit over 33%